Thread by @matt_barrie

I hope many read the Doherty modeling as it is both eye opening & huge fudge. The results rest on heroic contact tracing, indeed an addendum is added where presumable Doherty was told to go back & redo everything assuming "optimal" 'test trace isolate & quarantine'. 1/n #auspol

Taking the assumptions and modeling used at face value, at 50% vax we would need to go into lockdown harder than NSW is now 75% of the time (273 days/yr), 168 days at 60% and 80 days at 70%. You might say, well 80% is 0%! Well not really:

They end the simulations at 180 days, not even a full year or all of the seasons, and respiratory viruses are seasonal. You can see here how the simulation cuts off the data for the higher vaccination rates to make them look better. New infections don't even peak at 180 days...

Even so, the "let it rip" modeling shows daily new infections go to 55,000 (80% vax) - 70,000 (50% vax) even in the optimal “all adults” strategy. All the scenarios pretty much show us running out of COP ICU beds (1964 nationally, and other people get sick).

Again, the simulation time horizon has been fudged- 80% vax scenario doesn't even peak within the window. Even so, there are only 1964 ICU beds, and the health system is usually thinly funded & tightly balance. Vented patients take many more ICU bed days.

Deaths are also fudged due to the horizons. They don't even bother showing the 80% vax number it is fudged so badly. "World's best scientific analysis" says Scomo right now on the TV..

So this table is full of garbage data as it clips at 180 days and the graph above show how badly they clip.. even the 50% vax number clips.

They also use an r0 of 6.32 (not 8) when even their y-axis chart crime graphs which have been squished show r0 of 8.

r0 starts at 8 here, and I approximated the 50% and 80% scenarios with a linear scale.. basically its all in the "optimal test, contact tracing, isolate and quarantine".. which is already breaking down. 50% vax versus 80% vax is about 0.6 to 0.7 off the r0 budget with their data.

With realistic TTIQ, we need to stay into some form of lockdown under all the scenarios. Letting it rip will result in hard lockdown (harder than NSW in July 2021) 22-75% of the time by their own data.

Then, the entire paper is repeated as an "addendum" where it looks like Doherty has been told to go back and remodel everything because the results for "let it rip" don't look great at all! This is tacked onto the end after everything, so it looks like done at a later date.

The summary of the paper I will leave to Doherty themselves: Herd immunity is unlikely ‘even at very high levels of vaccine uptake (80% or above)’.

So this doesn't look like health advice, it looks like a fudged model for a political agenda. Everyone involved should now go away and read these papers (1) The Hammer and the Dance tomaspueyo.medium.com/coronavirus-th…

(2) COVID-19: A realistic approach to community management by Prof Robert Clancy.
quadrant.org.au/opinion/qed/20…

(3) COVID-19: Where Are We Going? by the same author.
quadrant.org.au/opinion/public…

What's a better approach that "letting it rip"? Eliminate the current outbreaks and build a proper quarantine to not let it in, that doesn't share ventilation, where workers are issued the proper equipment, outside of cities.

Here's the paper. I hope a few data scientists read it. doherty.edu.au/uploads/conten…

You've been given a gift here @AlboMP